IN THE SOUTH-WESTERN DISTRICT. 125 



tow-net eggs, on the outlying Eddystone rocks. Young examples are 

 common in the zostera beds of Cawsand Bay and the Yealm estuary, 

 but adults are rarely taken, at least by the Laboratory boats, on any- 

 ground fit for trawling. It would appear from our records that there is 

 no special migration in connection with the spawning instinct, but that 

 the ova are liberated on the grounds ordinarily inhabited by the species. 

 The breeding season appears to be prolonged from April to August, and 

 such measurements as were made afford evidence of a diminution in 

 the size of the ova as the season advances. The actual numbers will 

 be found in the records, while the subjoined list m.ay be taken as a 

 summary : — 



An egg of "67 mm., taken on the 28th June, can probably be assigned 

 to this species ; as it presents certain indications of immaturity, it 

 cannot fairly be utilised as evidence of size variation. Excluding this 

 specimen, the variation is "29 mm., or more than one-fourth of the size 

 of the largest specimen. 



It may be urged that the ova which we have assigned to C. nqjcstris 

 may really have been contributed by more than one species of wrasse, 

 but I do not think that this is the case. The common wrasses of 

 Plymouth are L. maculatus, L. mixtus, Cr. mclops, Ce. cxoletus, and Ct. 

 ruiKstris. The first three may be discarded, since their ova are 

 demersal. Of the ova of Centrolahrus cxoletus I know nothinf^, but I 

 found at Marseilles, where C. cxoletus is not known to exist, a similar 

 variation and seasonal diminution in the size of tow-net ova referable 

 to Ct. rupestris. 



April .... -80, -83 mm. 



May . . . . no observations. 



June . . . . -75, -76 mm. 



July . . , . -70 mm. 



Moreover the occurrence of young specimens in the Plymouth 

 zostera beds affords evidence that the spawning season is really as 

 prolonged as would appear from the tow-net gatherings. Preliminary 

 experiments indicate that the species can easily be reared from a very 

 small size, and it may be possible to study its development con- 

 tinuously. 



I 2 



